276 Extreme Response Style Bias in Burn Survivors: The LIBRE Project

Abstract

This study explores the extreme response style (ERS) of burn survivors who completed the Life Impact Burn Recovery Evaluation (LIBRE) Profile. The LIBRE Profile is an innovative patient reported outcome measure (PROM) that assesses social participation in burn survivors. Clinicians have recently begun to use PROMs to optimize care, however, PROMs can be sensitive to response style biases. ERS is a bias that occurs when an individual tends to select either the most positive or most negative response option of an item, when that choice does not reflect the person’s true ability or score. The objective of this study is to determine if the ERS distorts the LIBRE Profile scores. We fit a Multidimensional Generalized Practical Credit Model (MGPCM) with a high extreme response style (HERS) factor. We looked only at HERS because there were few participants who selected the low extreme response option. Next, we compared this model with the original MGPCM, which is not adjusted by the HERS bias, to estimate the impact that HERS has on scores. Lastly, we examined the personal characteristics, such as gender, age, time since burn, mode of administration and participation in peer support groups that may influence if an individual more likely to respond in a particular response style fashion would inflate their true low scores. The average impact of the HERS, based upon the root mean squared bias, ranged from 0.27–0.50 of a standard deviation of the scale. Individuals who were older, had participated in a burn survivor support group, and had selected to self-administer the measure were less likely to have a high HERS bias that masks low scores. Burn survivors displayed evidence of the HERS bias impacting their true LIBRE Profile scores. Burn survivors are a clinically unique group with special personal characteristics and clinical histories. In order to improve the clinical usefulness of the LIBRE Profile, measurement and understanding of extreme response styles is essential to aid interpretation of the scaled results. In future administrations, the HERS bias should be estimated to assess the occurrence of the bias and if it is affecting scores. Future work can consider this bias when measuring the psychosocial impacts of burn injuries and other health conditions.